UYAh AND WELLS
Coumermycin A I :A Preferential Inhibitor of Replicative DNA Synthesis in Escherichia coli. 11. In Vitro Characterization? Michael J. Ryan1 and Robert D. Wells*
ABSTRACT: In vitro inhibitions by coumermycin A I of DNA and RNA synthesis in toluenized cells were studied. In a sensitive strain, 50% inhibitions of replication and transcription were observed at 0.035 and 0.600 pg/ml, respectively. DNA synthesis in a toluenized-resistant mutant was 50% inhibited at 0.140 pg/ml of coumermycin A , , whereas RNA synthesis was unaffected at all concentrations tested. Studies with a mixture of toluenized-sensitive and -resistant bacteria ruled out the presence of a diffusable activator or inhibitor of coumermycin A I action. Density label studies with toluenized pol A+ and pol A- strains indicated that replicative DNA
synthesis was specifically inhibited, in agreement with the in vivo studies in the preceding paper of this issue (Ryan, M. J. (1 976), Biochemistry 1 5 ) . Highly purified Escherichia coli DNA polymerase 111 and RNA polymerase both were inhibited by this antibiotic. However, the high concentrations necessary for these inhibitions suggest that they are not biologically relevant. No interaction between DNA and coumermycin A , was observed with the following analytical procedures: ultraviolet difference spectra, DNA absorbance-temperature transitions, equilibrium buoyant density centrifugation, and DNA cross-linking determinations.
T h e preceding paper in this issue (Ryan, 1976) showed that coumermycin A I preferentially inhibited chromosomal DNA replication in vivo. Although in vivo studies can show a primary site of action of an antibiotic, they reveal little concerning its detailed mode of action. In order to characterize further the effects of this antibiotic on E. coli, in vitro experiments were performed, testing coumermycin A, against DNA and RNA synthesis in toluenized cells. In addition, the sensitivity of purified nucleic acid polymerases to coumermycin A I was determined and physical studies were performed to test the possibility of a direct interaction between this antibiotic and DNA. A portion of these results was reported previously (Ryan and Wells, 1975).
of ATP' and GTP, 100 pM each of UTP and CTP (with [3H]UTPlabeled at a specific activity of 40 to 70cpm/pmol). Assays were performed for 20 min at 36 "C and radioactive product was determined as described above. In almost all cases, assays were run in duplicate and the average values are reported. For DNA and RNA synthesis, each assay normally was within 5 or 10% of the average, respectively. E . coli RNA polymerase assays were as described (Burgess, 1969) except that 17 pM 4 x 1 7 4 replicative form DNA was used as template. Assays were performed at 30 OC with aliquots analyzed using procedure A (Ryan, 1976). E. coli DNA polymerase 111 was assayed at 26 OC essentially as described (Tamblyn and Wells, 1975) using as template 300 pM salmon sperm DNA that had been activated as described (Gass and Cozzarelli, 1973). Aliquots were analyzed using procedure A (Ryan, 1976). Drugs. Rifampicin was purchased from Sigma and streptolydigin was a gift to Julian Davies (this department) from the Upjohn Company. The latter two antibiotics were dissolved in Me2SO and stored at -20 OC. DNAs. XplacgS7CI8g7, E. coli, and 4X 174 replicative form DNAs were prepared and characterized as described (Chan and Wells, 1974; Wartell et al., 1974; Dodgson et al., 1976) and were gifts of I. Nes, J. E. Larson, and J. B. Dodgson, respectively. Reversal of Coumermycin A I Inhibition of E. coli R N A Polymerase. Two 0.1-ml reactions, one containing 50 pg/ml of coumermycin A1 and the second an equivalent amount of the drug diluent, MeZSO, were incubated at 30 OC. At regular intervals, three 0.02-ml aliquots were removed and, after 14 min, 0.02 ml of the reaction was diluted tenfold into an identical, prewarmed drug-free reaction mixture. Thereafter, 0.045-ml aliquots were taken at regular intervals. Other Materials and Methods. Other materials and techniques including drugs, enzymes, labeled and nonlabeled nu-
Materials and Methods Preparation of Toluenized Bacteria. Toluenization was carried out essentially as described (Moses and Richardson, 1970). Enzymes. E. coli DNA polymerase 111, prepared as described (Kornberg and Gefter, 1972), was a gift of T. Tamblyn and its characteristics were reported (Tamblyn and Wells, 1975). E. coli RNA polymerase was purified as described (Burgess, 1969), including the additional glycerol gradient centrifugation. This fraction was a gift of J. B. Dodgson. Assay Conditions. DNA synthesis in toluenized cells was assayed as described (Moses and Richardson, 1970). Reactions were r u n a t 36 "C for 45 min, and the entire assay was precipitated according to procedure B (Ryan, 1976). Assays for RNA synthesis in toluenized bacteria were as above except that the dNTP's were replaced by 500 pM each From the Department of Biochemistry, College of Agricultural and Life Sciences, University of Wisconsin, Madison, Wisconsin 53706. Receiced December 15. 1975. This work was supported by funds from the National Institutes of Health (CM-12275), the National Science Foundation (BMS74-21420), and a National Institutes of Health Training Grant (GM00236-14) to this Department. 3 Present address: Department of Chemistry, Massachusetts institute of Technolog), Cambridge, Mass. 021 39.
3778
BIOCHEMISTRY,
VOL.
15, NO. 17, 1916
'
Abbreviations used are: ATP, GTP, UTP, CTP, TTP, adenosine, guanosine, uridine. cytidine, and thymidine triphosphates; uv, ultraviolet.
IN VITRO INHIBITION OF D N A REPLICATION BY C O U M E R M Y C I N
A 4
ng. / ml FIGURE
A
A 0
A
0
A
1
Coumermycin A ,
1: Comparative inhibition of D N A synthesis by coumermycin
A,. Assays were performed as under Methods. Each 1.0-ml reaction contained 2 X los toluenized bacteria (0)D110;(A)CRBDl IO.
n g . / m l . Coumermycin A , 2: Comparative inhibition of R N A synthesis by coumermycin A,. Assays were performed as under Methods. The 0.4-ml reactions contained 8 X lo7 toluenized CRBDllO (A)or 1 X IO8 toluenized DI I O FIGURE
cleoside triphosphates, dBrUTP, thermal denaturation, and analytical sedimentation studies were as described (Ryan, 1976; Harwood et al., 1970; Grant et al., 1972; Burd et al., 1975; Wartell et al., 1975). Results Inhibition of Nucleic Acid Synthesis in Toluenized Bacteria. E . coli treated with toluene carry out ATP-dependent replicative DNA synthesis (Moses and Richardson, 1970) that continues from the same point at which the replicating fork stopped in vivo (Burger, 1971). When this system was tested using toluenized D110, it was found (Figure 1) that DNA synthesis was very sensitive to coumermycin A I . For the purpose of making quantitative comparisons, this reciprocal plot was useful, since there was a linear relationship between (product synthesized)-’ and the level of antibiotic in the assay. The amount of drug needed for a 50% inhibition of synthesis ( K ~ owas ) the absolute value of the intercept on the abscissa. There are a number of significant conclusions to be drawn from this experiment. First, since coumermycin Al-sensitive synthesis is a continuation of the in vivo process, it is clear that this antibiotic inhibits the elongation of chromosomal DNA. Second, since the necessary deoxyribonucleosidetriphosphates were supplied exogenously and inhibitions were still obtained, it can be concluded that this antibiotic did not function by inhibiting the synthesis of precursors. Third, assuming that there is a free exchange of both coumermycin A I and Mg2+between the assay medium and the interior of the cell, it is unlikely that this antibiotic acted simply by depletion of Mg2+,as suggested by other workers (Brock, 1967), since the concentration of Mg2+ is six orders of magnitude higher than the levels of drug used in these experiments. Finally, DNA synthesis in these cells was almost three orders of magnitude more sensitive than in vivo (Ryan, 1976) indicating a marked increase in cellular permeability. When a resistant derivative of this strain (CRBDllO) was toluenized and compared with its sensitive parent (D110) it was found (Figure 1) that it required 140 ng/ml of coum-
(0).
ermycin AI to inhibit its DNA synthesis 50%, a fourfold higher level than required for D110. When these two strains were toluenized and tested for sensitivity of their RNA synthesis (Figure 2) to coumermycin A I , it was found that approximately 600 ng/ml of coumermycin A I would have been required to give a 50% inhibition of this synthesis in D110, whereas its resistant derivative, CRBDl10, showed no detectable sensitivity even at the highest concentration tested (0.25 pg/ml). These results are in good agreement with the data obtained in vivo (Ryan, 1976), since (a) DNA synthesis was more sensitive to this antibiotic than RNA synthesis and (b) DNA synthesis in CRBDl 10 still showed some sensitivity to this drug but its RNA synthesis was not affected by the presence of coumermycin A,. These data suggest that DNA synthesis is the primary site of action of coumermycin A I especially since the growth rate of CRBDl10 is inhibited by this drug (Ryan, 1976) in the absence of any effects on transcription. Although toluenized D110 and CRBDl 10 demonstrated a differential sensitivity to coumermycin A,, they both were equally sensitive to nalidixic acid and rifampicin, specific inhibitors of DNA and RNA synthesis, respectively (Corcoran and Hahn, 1975). Nalidixic acid gave a 500h inhibition of DNA synthesis in both strains at 47 pg/ml, and rifampicin gave an equivalent inhibition of RNA synthesis at 38 ng/ml. Differential Inhibition of Replicative and Repair Synthesis of DNA by Coumermycin A I . In vivo studies indicate that coumermycin A I inhibited replicative DNA synthesis more strongly than repair synthesis (Ryan, 1976). A similar experiment was performed using toluenized 1 100 (poIA+ endI-) and D110 (polA- endI-). DNA synthesis in the latter strain was completely (95%) dependent on the presence of ATP, whereas it was not in the former strain. Each strain was incubated in reaction mixtures in which 5-bromodeoxyuridine triphosphate replaced dTTP, in the presence or absence of 100 BIOCHEMISTRY, VOL.
1 5 , N O . 17, 1976
3779
RYAN AND WELLS
To P
Bottom %
of
Gradient
Bottom
TOP
%
of
Gradient
F I G U R E 3: Buoyant density profile of 5-bromodeoxyuridylate-labeledD N A synthesized in the presence or absence of coumermycin A , , Reactions ( 2 ml each) were performed as described under Methods except that 30 fiM dTTP was replaced by 67 fiM dBrUTP, and ['HIdATP was added to a specific activity of 100 cpm/pmol. The assays were performed with two strains of toluenized E . coli: 1100 (polA+) and DI 10 (polA-). Cell lysis and CsCl density gradient centrifugation were performed a s described under Methods (Ryan, 1976). The arrow indicates the position of fully light 32PX DNA marker. Panel A is the profile obtained with toluenized 1100, panel B is that obtained with toluenized D1 IO. ( 0 )No drug, (0)100 ng/ml of coumermycin
AI.
ng/ml of coumermycin A ) . Buoyant density profiles of the DNA synthesized in each case are presented in Figure 3. Panel A shows that in thepolA+ strain, 1100, the DNA had two distinct peaks that corresponded to replicative and repair synthesis as defined previously (Ryan, 1976). After addition of coumermycin A I ,the repair peak was unchanged, whereas the hybrid density peak of replicative DNA synthesis was virtually abolished, indicating that this antibiotic had a clear specificity for replicative DNA synthesis in this system. The DNA product isolated from the polA- strain, D110, also had two peaks in the CsCl gradient (panel B). In this case, however, only about 20% of the products was repair synthesis. Addition of coumermycin A I to this reaction reduced the replicative DNA synthesis by apptoximately 95%. In this case (inset to panel B), there was more repair synthesis than replicative synthesis. This data indicates that repair synthesis of DNA was less inhibited than replicative synthesis; however, both were strongly inhibited in D110 even if to a somewhat varying degree. The repair synthesis seen in the polA+ strain, 1100, was likely the result of endogenous nucleases-generating primers that then were extended by DNA polymerase I. D110, the polA- strain, lacks this polymerase and, therefore, the repair synthesis seen with this strain may employ factors that were shared with a replicative DNA synthetic complex, e.g., DNA polymerases I1 and 111 (Tait et al., 1974). If this were the case, it might explain why this repair synthesis was inhibited by coumermycin A I . Thus, it is concluded that replicative synthesis was specifically inhibited by coumermycin A , , whereas repair synthesis, at least that mediated by the E . coli DNA polymerase I , was not affected. Irreuersibility of Inhibition of D N A Synthesis by Coumermycin A , in Toluenized Cells. Attempts were made to re-
3780
BIOCHEMISTRY, VOL.
15,
NO.
17, 1976
verse the inhibition of DNA synthesis in toluenized D110 by diluting the reaction tenfold into a drug-free assay mixture. Alternatively, the cells were centrifuged after a short incubation with the drug and were resuspended in an equal volume of drug-free reaction mixture (data not shown). Both experiments indicated that, within the time periods examined, coumermycin A, inhibition of DNA synthesis was not reversed. This could indicate the formation of a drug-receptor complex that was very stable, once formed under these conditions. Absence of a Diffusable Activator or Inhibitor of Drug Action. The observed differential sensitivity of D110 and CRBDllO to coumermycin A, might be explained by the presence of an activator or inhibitor of drug action present in the toluenized bacteria. Testing this possibility involved performing reactions with each strain alone in the presence and absence of coumermycin A I . This data was then compared with that obtained with a mixture of both strains of toluenized bacteria. Table I shows that the DNA synthesis, both in the presence and absence of coumermycin A,, of each strain appeared to be independent of the other, since the incorporation for the mixture of strains was equal to the sum obtained with each assayed separately. The close agreement between the sum of 1 and 2 and the actual incorporations (3) in each case indicated the absence of diffusable factor that either potentiated or diminished the inhibitory action of coumermycin A1 in this system. Absence of a Detectable Interaction between Coumermycin A I and DNA. Since most of the data obtained could be explained by the presence of a drug-DNA interaction, studies were performed to test for such a complex. Two determinations of the ability of the drug to perturb DNA properties and one determination of the ability of DNA to perturb the drug properties were performed. Thermal denaturation studies (data not shown) were carried out with E . coli DNA. When added,
IN VITRO INHIBITION O F D N A REPLICATION B Y COUMERMYCIN TABLE I: Additivity of D N A Synthesis in Toluenized D110 and
C R B D l 10 in the Absence and Presence of Coumermycin A , . a
.
T pmol of [ c Y - ~ * P ] ~ G M Incorporated P
g
Toluenized Strain
-
(1)CRBDllO (2) D110 Sum of (1) (2) performed separately (3) Mixture of C R B D l l O and DI 10
+
Coumer- Coumermycin AI mycin A I Control (+55 ng/ ( + I 1 0 (+ diluent) ml) ng/ml) ~
~~~
B
4
c
8
3
a
I
V
~
v
454
76 85 161
111
457
173
119
129
325
55 56
2
r Y
a
Assays were performed as described under Methods. Three sets of 0.9-ml reactions were carried out with the first containing 1 X IOs toluenized C R B D l 10, the second containing 3 X IO8 toluenized D110, and the third containing both 1 X 10s,toluenized C R B D l 10 and 3 X IO8 toluenized D110. Each set was made up of three reactions, the first containing no drug, and the second and third containing 55 and 110 ng/ml of coumermycin A I , respectively.
IO
20
30
40
Time ( m i n )
Inhibition of E. coli D N A polymerase I11 and E. coli R N A polymerase by coumermycin A I . Assay conditions were as described under Methods. Inhibition of DNA polymerase 111 and R N A polymerase are shown in panels A and B, respectively. ( 0 )N o drug, (0)10 @/mi of coumermycin A I , (0)20 pg/ml of coumermycin A I , (X) 50 pg/ml of coumermycin A I , (m) 10 pg/ml of streptolydigin. F I G U R E 4:
W
ul
coumermycin A I was present in the ratio of 1 drug molecule/ base pair of DNA. The Tm'swere identical in the presence and absence of coumermycin A I and were 72 OC (14 m M Na+) and 86 OC (0.1 M Na+). The analytical buoyant density of E. coli DNA in the presence and absence of coumermycin A I was also tested as described (Wartell et a]., 1974); no change was detected at the ratio of 2 drug molecules/bNA base pair. Coumermycin A I absorbs strongly in the uv region with a EM of approximately 6 X lo4 at 282-284 nm. Since a drugmacromolecule interaction might be expected to alter this spectrum, difference spectra were studied. No effect of X DNA on the spectrum of coumermycin A1 was observed at the ratio of 1 drug niolecule/l.6 base pairs. However, bovine serum albumin, present at 69 yglml, had a considerable influence on the spectrum of coumermycih A I , inducing a hypochromic shift below approximately 3 18 nm (1 1% hypochromism at 285 nm) and a slight hyperchromic shift at longer wavelengths (data not shown). From this experiment it is suggested that there is an interaction, probably nonspecific, between coumermycin A I and bovine serum albumin. However, there is no indication of an interaction between this antibiotic and DNA. Inhibition of Purified Polymerases by Coumermycin A I . E. coli DNA polymerase 111 and R N A polymerase are essential for the in vivo processes of replication and transcription, respectively, and, therefore, are possible target sites for coumermycin A I . When tested, both were 50% inhibited by coumermyin A I at levels between 10 and 20 pg/ml (Figure 4). These values are approximately 30- and 200-fold higher than the levels required for inhibition of RNA and DNA synthesis, respectively, in toluenized D110. In contrast, streptolydigin gave a 50% inhibition of R N A synthesis in toluenized cells at 5-10 pg/ml (data not shown), and approximately an 80% inhibition of the purified RNA polymerase at 10 wg/ml. Despite the apparent sensitivity of these purified enzymes t o coumermycin A1, it is not likely that they are specific inhibitions. For example, in contrast with the results obtained in toluenized cells, a tenfold dilution of an R N A polymerase assay immediately and completely reversed a total inhibition of this enzyme by coumermycin A I (Figure 5). There is also a considerable amount of evidence that this antibiotic can interact
0 L
-E W
)r
4
0
0
4
-
a" A
/
Time ( m i n ) 5 : Reversal of coumermycin A I inhibition of R N A polymerase. Assays were performed as described under Methods, either in the presence (0)or absence (A)of 50 yg/ml of coumermycin A I . After 14 min of incubation (arrow), each reaction was diluted tenfold, with the concentration of cohmermycin A I reduced to 5 pg/ml (x) in one assay. FIGURE
nonspecifically with proteins including: ( 1 ) the difference spectrum obtained above with bovine serum albumin; (2) the observation (Fedorko et al., 1969) that the presence of serum proteins raises the minimal inhibitory concentration for coumermycin A I from 0.004 yg/ml to 6.25 yg/ml, a 1500-fold increase, when tested against strains of S . aureus; (3) the large number of inhibitions of unrelated enzymatic activities reported for novobiocin (Brock, 1967), an antibiotic closely related to coumermycin AI. In summary, it is concluded that the inhibitions of E. coli DNA polymerase I11 and RNA polymerase are probably nonspecific and unrelated to the in vivo mechanism of inhibition by this antibiotic. Discussion Coumermycin A l at very low concentrations preferentially inhibits DNA replication in toluenized E. coli. However, it is BIOCHEMISTRY, VOL.
15, NO. 17, 1976
3781
R Y A h
somewhat perplexing that, in vitro, transcription is so much less sensitive than replication to coumermycin A , compared to the results obtained in vivo (Ryan, 1976). These observations are in agreement with effects seen with novobiocin in which DNA synthesis was inhibited to a greatei degree than RNA synthesis in vivo (Smith and Davis, 1967). However, in vitro studies with toluenized cells (Staudenbauer, 1975) demonstrated an 80% inhibition of replication at 2 kg/ml of novobiocin with no apparent affect on transcription. Thus, in vitro coumermycin A I and novobiocin are specific inhibitors of DNA replication, whereas in vivo the differential effects on replication and transcription are much less pronounced. These observations have led us to refer to coumermycin A , as a “preferential” as opposed to a “specific” inhibitor. One hypothesis to explain this is that, in vivo, there is some control mechanism that represses transcription when replication is inhibited and this mechanism is not functional intoluenized cells. However, there is no evidence for this and, in fact, it is contrary to observations with nalidixic acid (Goss et al., 1965) and 6-(p-hydroxyphenylaza)uracil (Brown and Handschumacher, 1966). In both cases, almost total inhibition of DNA synthesis was obtained with negligible effects on RNA synthesis. Furthermore, there was no inhibition of uv-induced repair synthesis of DNA by 6-(p-hydroxyphenylaza)uracil despite a complete inhibition of DNA replication (Brown, 1971). Another possibility is that RNA synthesis in toluenized cells is not representative of in vivo transcription. However, this is not likely, since this system was tested by others and found to correspond well with known properties of transcription (Peterson et al., 1971). A third possibility is an extension of a speculative model (Ryan, 1976) in which coumermycin A, inhibits transcription and replication by forming a highly stable complex with a target protein and the chromosomal DNA. The synthesis of RNA in vivo could be less sensitive to this complex because it would have a greater ability to bypass or overcome this block than does DNA synthesis. This was proposed since, in drug resistant mutants, RNA synthesis was unaffected by coumermycin A , . whereas DNA synthesis still showed a reduced sensitivity to this antibiotic. Since the internal environment of the cell after toluenization is probably different from that in vivo, it could be hypothesized that there is a slight reduction of the stability of the inhibitory complex that is sufficient to allow transcription to proceed while replication is still inhibited. The observation that there is an apparent interaction between coumermycin A! and bovine serum albumin, in addition to other studies (Brand and Toribara, 1975) with the closely related novobiocin, suggests that these antibiotics are capable of nonspecific interactions with proteins, perhaps by a hydrophobic interaction. The possibility of such interactions is important i n evaluating the inhibitions of in vitro polymerase reactions, since the concentration of drug used is critical. Our results with toluenized E . coli, combined with the minimal inhibitory concentrations for other organisms (Fedorko et al., I969), indicate that potent inhibitions can be found at levels of 0.04 to 0.004 pg/ml. Thus, inhibitions of purified DNA and R h A polymerases by 10-20 pg/ml of coumermycin A I are probably nonspecific. The mode of action of coumermycin A ,
3782
BIOCHEMISTRY,
VOL.
15, N O . 1 7 , 1976
A N D WtL1.S
will be established with certainty only when it can be shown that a purported target site, isolated from a drug-resistant bacterium, confers resistance on an in vitro, reconstituted system for the replicative synthesis of DNA.
Acknowledgment We thank Jacquelynn Larson for performing the analytical sedimentation experiments. References Brand, J. G., and Toribara, T. Y. (1975), Arch. Biockem. Biophys. 167, 91. Brock, T. D. (1967), Antibiotics 1, 651. Brown, N . C . ( I 97 l ) , J . Mol. Biol. 59, 1. Brown, N. C., and Handschumacher, R. E. (1966), J. Biol. Chem. 241, 3083. Burd, J. F., Wartell, R. M., and Wells, R. D. (1 975), J. Biol. Chem. 250, 5 109. Burger, R. M. (1971), Proc. Natl. Acad. Sci. U.S.A. 68, 2 124. Burgess, R. R. (1969), J . Biol. Chem. 244, 6160. Chan, H. W., and Wells, R. D. ( 1 974), Nature (London)252, 205. Corcoran, J. W., and Hahn, F. E. (1975), Antibiotics 3, Dodgson, J . B., Nes, I. F., Porter, B. W., and Wells, R. D. (1976), Virology 69, 782. Fedorko, J., Katz, S.,and Allnoch, H. (1969), Appl. Microbiol. 18. 869. Gale, E. F., Cundliffe, E., Reynolds, P. E., Richmond, M. H., and Waring, M. J . (1972), The Molecular Basis of Antibiotic Action, London, England, Wiley. Crass, K. B., and Cozzarelli, N. R. (1973), J . B i d . C‘hem. 248, 7688. Goss, W. A., Dietz, W. H., and Cook, T. M. (1965). J . Bacteriol. 89, 1068. Grant, R. C., Kodama, M., and Wells, R. D. (1972), Biochemistry 1 1 , 805. Harwood, S. J., Schendel, P. F., and Wells, R. D. (1970), J . B i d . Chem. 248, 561 4. Kornberg. T., and Gefter, M. L. (1972), J . Biol.Chem. 247, 5369. Moses, R.E..and Richardson, C. C. (1 970), Proc. Natl. Acad. Sei. U.S.A. 67, 674. Peterson, R. L., Radcliffe, C. W., and Pace, N . R. (1971), J . Bacteriol. 107, 585. Ryan, M .J. (1976), Biochemistry I 5 (preceding paper in this issue). Ryan, M. J., and Wells, R. D. (1975), Fed. Proc. Fed. A m . SOC.E x p . Biol. 34, 1896. Smith, D. H., and Davis, B. D. ( 1 967), J . Bacteriol. 93, 7 1 . Staudenbauer, W. L. (1975), J . Mol. Biol.96, 201. Tait, R. C., Harris, A. L., and Smith, D. W. (1974), Proc. .2’atl. Acad. Sci. U.S.A.71, 625. Tamblyn, T. M., and Wells, R. D. (1975), Biochemistry 14, 1412. Wartell, R. M., Larson, J. E., and Wells, R. D. (1974), J. Biol. Chem. 249, 6719. Wartell. R. M., Larson, J. E., and Wells, R. D. (1975), J . Biol. Chem. 250, 2698.
